Ben Tal Mizrahi Tel Aviv
I work at the edge where customers meet hard infrastructure problems. At Fireblocks, I take ambiguous custody, wallet, and settlement issues, narrow them into clear technical paths, then build the systems and automation that make those paths repeatable.
Four years of customer-facing technical work across digital asset custody, wallets, off-exchange settlement, and integrations. I am comfortable in constrained, security-sensitive environments where access is limited and a wrong move is expensive.
The work I'm proudest of is TRBT, a multi-agent system I built at Fireblocks that a support team now uses every day to investigate customer issues faster and with better evidence.
Receipts, not vanity metrics
Real customer context in high-stakes crypto infra, automation shipped and in daily use, and a track record of bridging technical and commercial teams without losing either side.
TRBT, Tickets Resolved Before Tea, is a multi-agent AI system I built at Fireblocks. It reads a complex request, picks the right specialist agents, checks trusted data and logs, then turns the findings into a report an engineer can act on.
Field notes from the customer edge: signing flows under load, reverted contracts, and constrained provider mismatches. Each problem became a customer-ready answer or a reusable investigation pattern.
Read the field notes →A tool I built to do the investigation work faster and more consistently. An orchestrator and around 20 focal agents that check trusted data sources and return a cited report a human can act on.
Read the case study →Proof that the build instinct extends into production: customer reporting workflows from idea to live delivery, with repeatable runs and clearer handoffs. n8n, a SQL data warehouse, LLM summaries, practical PM judgment.
See the breakdown →Work close to customers, scope integrations, debug real deployments, and build the tooling that unblocks them. Discovery, demos, integration scoping, security reviews. Digital asset custody, wallets, off-exchange, fintech.
Trusted technical voice with prospects and customers: discovery, demos, RFPs, security questionnaires, and translating product depth into a path the buyer can act on.
Own the workflow, model selection, integrations, guardrails, observability, and QA for AI systems teams can depend on. Measure time saved, handoff quality, and whether teams keep using it.
Fintech, web3, payments, and digital asset custody are the domains I've spent the last four years inside. APAC and EMEA both work. If the seat you're hiring for sits on that line, say hi.
Hiring for a role where customer-facing engineering meets AI systems? Building AI workflows for a technical GTM or product team? Same answer:
